1. Field of the Invention
The present invention relates to an image capture apparatus and a method of controlling the same, and in particular relates to an automatic focus detection technique.
2. Description of the Related Art
As a technique for realizing automatic focus detection (AF) that employs a phase-difference detection method using signals obtained from an image sensor, there is known to be a technique of giving a specific function to some of the pixels arranged in the image sensor. For example, Japanese Patent Laid-Open No. 2000-292686 discloses an image sensor in which pixels for focus-detection, which have been given a pupil-dividing function by arranging the light-receiving area so as to be shifted relative to the optical axis of the on-chip microlens, are arranged inside the pixel region at a predetermined interval. Using this type of image sensor makes it possible to perform focus detection employing a phase-difference detection method based on a pair of signals obtained from pixels for focus-detection that have different pupil division directions.
Meanwhile, image capture apparatuses of recent years have employed a method suited to signal application when reading out signals from the pixels of an image sensor. For example, in the case of performing live-view display, since the resolution of the display apparatus is lower than the pixel count of the image sensor, a signal corresponding to a pixel count suited to this display is obtained by adding up multiple pixels during readout or thinning out pixels during readout. When recording a moving image as well, images having a moving image resolution are acquired in a similar manner. Capturing and recording a moving image for live-view display differs from still image recording in that there is desire for a high frame rate since the smoothness of the display is more important than the resolution.
However, if the frame rate is raised, the exposure time per frame decreases. It is therefore difficult to maintain precision in automatic focus detection that uses signals obtained from pixels for focus-detection, particularly in the case of a low-brightness subject. In order to address this issue, Japanese Patent Laid-Open No. 2008-085738 proposes a method of adding up multiple frames worth of signals obtained from the pixels for focus-detection when performing focus detection.
However, Japanese Patent Laid-Open No. 2008-085738 does not disclose how focus detection is performed in the case of driving the focusing lens during moving image capture. If the subject image changes due to the driving of the focusing lens, the change in the subject image will influence the image misalignment amount of the added-up image signals. For this reason, there are cases where the in-focus position is not reached even when performing relative driving of the focusing lens based on the focal position at the center time of the accumulation period of the added-up frames.